ISSN: 1941-3084 Copyright © 2009 American Heart Association. All rights reserved. Print ISSN: 1941-3149. Online

Greenville Avenue, Dallas, TX 72514Circulation: Arrhythmia and Electrophysiology is published by the American Heart Association. 7272

DOI: 10.1161/CIRCEP.108.778357 2009;

2009;2;258-267; originally published online Mar 6,Circ Arrhythm ElectrophysiolJens Broscheit, Peter M. Jakob and Oliver Ritter

Hiller, Matthias Nahrendorf, Michelle Maxfield, Sabine Wurtz, Wolfgang Geistert, Peter Nordbeck, Wolfgang R. Bauer, Florian Fidler, Marcus Warmuth, Karl-Heinz

ElectrophysiologyFeasibility of Real-Time MRI With a Novel Carbon Catheter for Interventional

http://circep.ahajournals.org/cgi/content/full/2/3/258located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://www.lww.com/reprintsReprints: Information about reprints can be found online at  

journalpermissions@lww.com410-528-8550. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters 

http://circep.ahajournals.org/subscriptions/online at Subscriptions: Information about subscribing to Circulation: Arrhythmia and Electrophysiology is

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

Feasibility of Real-Time MRI With a Novel CarbonCatheter for Interventional Electrophysiology

Peter Nordbeck, MD; Wolfgang R. Bauer, MD, PhD; Florian Fidler, PhD;Marcus Warmuth, BSc; Karl-Heinz Hiller, PhD; Matthias Nahrendorf, MD; Michelle Maxfield, BSc;

Sabine Wurtz, MSc; Wolfgang Geistert, PhD; Jens Broscheit, MD; Peter M. Jakob, PhD; Oliver Ritter, MD

Background—Cardiac MRI offers 3D real-time imaging with unsurpassed soft tissue contrast without x-ray exposure.To minimize safety concerns and imaging artifacts in MR-guided interventional electrophysiology (EP), we aimedat developing a setup including catheters for ablation therapy based on carbon technology.

Methods and Results—The setup, including a steerable carbon catheter, was tested for safety, image distortion,and feasibility of diagnostic EP studies and radiofrequency ablation at 1.5 T. MRI was performed in 3different 1.5-T whole-body scanners using various receive coils and pulse sequences. To assess unintentionalheating of the catheters by radiofrequency pulses of the MR scanner in vitro, a fluoroptic thermometry systemwas used to record heating at the catheter tip. Programmed stimulation and ablation therapy was performed in 8pigs. There was no significant heating of the carbon catheters while using short, repetitive radiofrequency pulsesfrom the MR system. Because there was no image distortion when using the carbon catheters, exact targeting ofthe lesion sites was possible. Both atrial and ventricular radiofrequency ablation procedures including atrioven-tricular node modulation were performed successfully in the scanner. Potential complications such as pericardialeffusion after intentional perforation of the right ventricular free wall during ablation could be monitored in realtime as well.

Conclusion—We describe a newly developed EP technology for interventional electrophysiology based on carboncatheters. The feasibility of this approach was demonstrated by safety testing and performing EP studies and ablation therapywith carbon catheters in the MRI environment. (Circ Arrhythmia Electrophysiol. 2009;2:258-267.)

Key Words: ablation � electrophysiology � MRI

Although interventional therapy of complex arrhythmiasis becoming more and more common, these approaches

are still accompanied by a substantial risk of proceduralcomplications.1 Therefore, real-time imaging of the surround-ing anatomy, lesion mapping, and early detection of compli-cations seem to be crucial for the further development ofthese interventional approaches.

Clinical Perspective on p 267In this context, MRI has been proposed as an alternative

imaging modality for guiding and monitoring electrophys-iological (EP) procedures.2 MR technology combines var-ious strengths—providing excellent 3D information onanatomy, cardiac function and size, and the location ofablation lesions—with the benefit of examinations withoutionizing radiation or iodized contrast agents. Furthermore,

an assessment of the extent of the ablation lesions andidentification of possible remaining gaps could be possiblein the future.3,4

However, to perform EP interventions with MRI guidance,a special setup must be established, and the catheters usedmust fulfill many requirements. There are concerns related tomagnetic field–induced movement and substantial radiofre-quency (RF)-induced heating at device-to-tissue interfaces,potentially resulting in thermal tissue damage around the lead.The electrically conductive lead may pick up electromagneticinterference from the MR system, which can translate into heator produce a life-threatening tachycardia in the patient.5–7 Ad-ditionally, image distortion by the metallic intracardiac lead canreduce the advantage of high-resolution imaging.8

MRI-compatible devices and strategies would open newavenues for treatment options in the field of invasive

Received March 10, 2008; accepted February 20, 2009.From the Department of Internal Medicine I (P.N., W.R.B., O.R.), the Department of Experimental Physics V (P.N., M.W., P.M.J.), and the Department

of Anesthesiology (J.B.), University of Wurzburg, Germany; Research Center Magnetic-Resonance-Bavaria (F.F., K.-H.H.), Wurzburg, Germany; Centerfor Molecular Imaging Research (M.N.), Massachusetts General Hospital, Harvard Medical School, Boston, Mass; and Biotronik GmbH & Co. KG(M.M., S.W., W.G.), Berlin, Germany.

Drs Nordbeck and Bauer contributed equally to this work.The online-only Data Supplement is available at http://circep.ahajournals.org/cgi/content/full/CIRCEP.108.778357/DC1.Correspondence to Oliver Ritter, MD, Department of Internal Medicine I, University of Wurzburg, Josef Schneider Str. 2, 97080 Wurzburg, Germany.

E-mail Ritter_o@klinik.uni-wuerzburg.de© 2009 American Heart Association, Inc.

Circ Arrhythmia Electrophysiol is available at http://circep.ahajournals.org DOI: 10.1161/CIRCEP.108.778357

258 at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

electrophysiology. The objectives of this study were (1)developing carbon catheters and a compatible setup forinterventional EP in the MR environment, (2) subsequentdemonstration of the feasibility of diagnostic EP proce-dures, and (3) ablation therapy under real-time MRIguidance.

Methods

Catheter DesignA custom EP ablation catheter with carbon fiber conductors wasdesigned for use in the MRI environment. In this steerable catheter,unlike standard EP ablation catheters, the steel elements such aswires, coils, crimping hubs, and bearing surfaces were replaced byplastic or ceramic alternatives. This was done to eliminate potentialimaging artifacts and to prevent interactions with the static field. Thecatheter body was made of 7F Pebax tubing reinforced withnonconductive fibers and a 7F steerable tip. The catheter’s usablelength was approximately 85 cm, and the 8 cm steerable tip could bedeflected from straight position to 120° to achieve precise electrodeplacement. The large carbon fiber bundle (approximately 30 000fibers) used for the distal electrode yielded a low resistance of 3.8ohms, whereas the smaller carbon bundle (approximately 3000fibers) had a higher resistance of 126 ohms. The platinum/iridiumelectrodes were connected to the carbon bundles by means ofsilver conductive epoxy, crimp rings, and copper connectingwires soldered to the electrodes. The catheter had a 4-mm-longablation electrode, a 2-mm-long ring electrode, and 2-mm spac-ing. These carbon-based circuits were designed to allow forhigh-energy RF ablations through the distal electrode and low-energy sensing/pacing through the bipolar pair made from thedistal and proximal electrodes. The catheter was connected to theRF generator and to the external pacer (ERA 300, BiotronikGmbH, Berlin, Germany) through customized long cables and afilter box that allowed the equipment to remain in the scannerroom. A custom-made neutral electrode with a large skin contactpatch (100�200 mm) and a carbon lead was used for RF deliveryto avoid adverse effects in the MRI environment.

MR SystemsMR examinations were performed on 3 different 1.5-T whole-body scanners: a Magnetom Vision (hereafter referred to asScanner 1), a Magnetom Avanto (Scanner 2, both SiemensMedical Solutions, Erlangen, Germany), and a Gyroscan ACS NTIntera R12 (Scanner 3, Philips Medical, Best, The Netherlands),using the integrated body coil for radiofrequency excitation.Signal reception was performed using either the body coil oradditional receiver coils, depending on the scanner and experi-ment. For in vivo MRI experiments on pigs, fast, non–ECG-triggered scout images (ie, using fast low-angle shot [FLASH]pulse sequences) were used to position the imaging slice in theregion of interest, for example, the midventricular short-axisplane of the heart. Cine movies and high-resolution images ofspecific heart phases were acquired using an ECG trigger module.Detailed descriptions on the various pulse sequences implementedon the 3 scanners can be found below.

MR Compatibility/Safety TestingFor testing of heat development caused by RF pickup during MRI,various in vitro investigations of conventional (metallic) andcustom-made (carbon) EP catheters were performed in Scanners 1and 2. In Scanner 1, heating tests of the catheters were performedin a gelatin-casted block containing a fresh pig heart. The gelatincontained 0.9% NaCl solution and the size of the phantom was300�230�130 mm. This phantom was centered inside thetransmit coil. RF-related catheter heating during MRI was eval-uated by temperature measurements using an external fluoropticthermometry system (Labkit m3300 with connected STF-10

probes, Luxtron, Santa Clara, Calif). Temperature probes wereattached to different segments of the catheters, and the tempera-ture distribution was continuously recorded while performingMRI.

Additional heating tests were performed in Scanner 2 within agel-filled head and torso phantom as previously described.9 Waterdoped with 0.15% NaCl and 3% hydroxyethylcellulose (volume, 45L) was used as phantom filling, providing conductivity valuescomparable to that of body tissue (0.47 S/m) while preventingconvective heat transport. Ten representative catheter configurationswere tested for each catheter, including straight, curved, half circle,and full circle configurations, with the catheter always crossing fromair to gel at the lower right phantom side. All investigations wereperformed with temperature probes at the tip, 200 mm behind the tip,at the crossing from air to gel, and apart from the catheter in the gel(reference). Catheter positions and orientations were tested forRF-related heating using the fluoroptic thermometry system whilerunning an SSFP imaging sequence, with the following imagingparameters: echo time, 1.65 ms; repetition time, 3.29 ms; bandwidth,543; field of view, 500 mm; slice thickness, 10 mm; matrix, 64�64.Total acquisition time was set to 5 minutes. Transmit power of theMR system was adjusted by variation of the flip angle between 2 Wand 131 W, relating to a specific absorption rate between �0.1 and2.8 W/kg.

In addition to the experiments on MRI safety, in vitro investiga-tions of the custom-made catheter on MRI compatibility wereperformed at 1.5 T in the phantoms described above to prepare forthe in vivo experiments. Various pulse sequences were tested fordetermination of possible artifacts and visualization properties of thecatheter.

In Vivo Imaging ProtocolAs a result of the safety and compatibility investigations, shortFLASH and TrueFISP (echo time, 1.35 ms; repetition time, 2.69 ms;flip angle, 80°; field of view, 400 mm; slice thickness, 8 mm; matrix,168�256; total acquisition time, 0.7 s) imaging sequences for in vivocatheter tracking purposes were developed. These sequences wereadapted for low RF transmit power/SAR (lowest value still account-ing for admissible image quality �4 W��0.1 W/kg), while en-abling detection of the carbon catheters in vivo and guidance to theablation sites as demonstrated later.

Experimental Setup for Animal ExperimentsAll animal protocols were approved by the governmental animal careand use committee (Regierung von Unterfranken, approval No. 54 to2531.01 to 63/04). One dead minipig was used to establish the setup;a total of 8 minipigs weighing 30 to 54 kg were then sedated with a10-mg intramuscular injection of ketamine and maintained on 1%isoflurane at the initial preparation of the jugular veins and placingof the catheters. Further experiments were performed under intrave-nous sedation/analgesia (midazolam/fentanyl/rocuroniumbromide).For ventilation during the MR procedures, an Oxylog ventilator(Draeger Medical, Luebeck, Germany) was located outside thescanner room and the animals were ventilated using a 4-m extensiontube. 8F introducer sheaths were placed in the jugular veins forcatheter access. The pigs were positioned supine, feet first, inside thescanner, allowing access to the sheaths. Signal reception in Scanner1 was performed using a 4-element body array. In Scanner 2, varioussetups were tested, including signal reception with the body coil, amultiple flex coil assembly, and an experimental 16-channel surfacecoil. In Scanner 3, a 5-channel cardiac array was used. A broadspectrum of diagnostic and interventional EP procedures was per-formed in the animals. Detailed descriptions on some of the testcomponents are given below.

EP/Ablation ProtocolFor diagnostic EP studies including programmed stimulation, differ-ent anatomic regions were targeted in each animal including the rightatrium (lateral free wall), right ventricular apex, the His bundle

Nordbeck et al Hard- and Software Advances for MRI-Guided Ablation Therapy 259

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

region, and coronary sinus. Pacing and sensing thresholds wererecorded in each region.

After systematically directing the carbon catheter to the 4 ana-tomic regions in each animal, we proceeded to an ablation protocol,aiming to demonstrate that successful RF energy application ispossible in 3 different MR scanners using the carbon catheters duringMRI to track the tip of the catheter. An EPT-100XP generator(Boston Scientific Corporation, San Jose, Calif) was located outsidethe scanner room. The neutral electrode and the carbon catheterswere connected to a receiver for sensing and an external paceroutside the scanner room for pacing using 6-m extension cables. Amajor limitation of energy delivery in terms of ablation and acqui-sition of IEGMs in the MR scanner is electromagnetic interference.Although the frequency (350 kHz) of the RF generation unit(EPT-100XP) is significantly lower than the 64-MHz proton preces-sion frequency at 1.5 T, the rectangular-shaped pulse form of theablation generator implied higher harmonics interfering with imageacquisition, which regularly results in signal void during imaging.Therefore, a special RF filter was used to suppress these higherharmonic signals. This low-pass filter consists of diverse electriccomponents cutting off interfering frequencies above 5 MHz. Thecutoff frequency for the HF poles (catheter tip and skin patch forablation) is above 5 MHz and for the ECG poles (ring electrode ofthe catheter) above 1 kHz, for the temperature channels above 5 kHz.RF energy from the ablation generator is directed through the filterbox to the ablation catheter (complementary Figure 1). A similar

approach has been described earlier.2,10 IEGM tracings were ac-quired via the carbon catheters and directed outside the scanner roomthrough the filter system. Recordings were performed with auto-mated data acquisition software (LabSystem Duo, BARD Electro-physiology, Lowell, Mass).

One ablation target was chosen in each animal, including sites inthe right atrium, right ventricle, atrioventricular (AV) node, andcoronary sinus, to simulate different interventional EP approaches.Ablation was performed in cycles for 60 seconds with maximumpower of 75 W and temperature cutoff at 65° C. After baselineimages were acquired for the optimal slice selection, RF ablation wasperformed at the site between the distal electrodes and the customneutral electrode patch, which was adhered to the skin of the lowerabdomen; ideal contact was secured using commercial contact gel.The primary end point of RF delivery was a significant reduction(�80%) of IEGM voltage. RF delivery was interrupted whenbradycardia or tachycardia occurred. Although lesion mapping wasbeyond the scope of this study, we assessed lesion size during theprocedure to gain intraprocedural information of whether there wassuccessful creation of lesions. For this purpose, gadolinium-enhanced MRI was used according to previously publishedprotocols.3

After the MR experiments, the animals were euthanized byintravenous administration of T61 (Intervet, Unterschleisheim, Ger-many): 1 mL contains Tetracain-HCL 5.00 mg; Mebezoniumiodid50.00 mg; Embutramid 200.00 mg. The hearts were excised andsectioned to identify the thermal lesions. Extent and location of therespective lesion was recorded by gross examination.

Statistical AnalysisChanges in pacing and sensing threshold levels were analyzed duringthe time course and were considered significant at a level of P�0.05in a paired t test (ANOVA). Comparisons of ablation lesions in MRIand in pathological specimens were performed by a paired t test(ANOVA). Unless noted otherwise, results measurements are re-ported as arithmetic mean�SD. Statistical analysis was performedusing SPSS 14.0.1.

Results

Testing for Unintended HeatingIn this work, custom-made carbon EP catheters (Figure 1A)were compared with commercially available conventionalmetallic ablation catheters (EPT Blazer, Boston ScientificCorporation, San Jose, Calif) regarding MR safety/compati-bility. Metallic and carbon catheters were tested in variousphantom configurations (Figure 1B). Both of the catheterpathways in the gel and in the air strongly affected themagnitude of tip heating. The metallic catheters showed thehighest MR-related temperature elevation (catheter tip heat-ing) in a configuration as shown in Figure 1B. Whenperforming SSFP imaging with a total acquisition time of 5minutes and an SAR of 2.1 W/kg, the temperature at themetallic catheter tip increased from 17.3�0.4° C to a maxi-mum of 63.5�0.2° C; maximum heating at the carboncatheter tip in all configurations (SAR 2.1 W/kg, 5 minutes)was an increase of 13°C (Figure 2). In a given configuration,heating was generally found to be proportional to the RFpower applied by the MR system (Figure 2B), which is in linewith previous investigations.9 Average temperature increaseat the catheter tip was 22�4.1°C in metallic catheters and4.0�1.3°C in carbon catheters (n�10; P�0.05) (Figure 2C).At the point where the catheter crossed from air into gel,mean heating was 2.5°C at the conventional and 2.3°C at the

Figure 1. A, Original photograph of a carbon-based EP cathe-ter: close-up showing the grip and the catheter tip. Thedesign resembles conventional EP catheters. The catheter issteerable. B, Original photograph showing the gel phantomwith the carbon catheter for testing of temperature develop-ment during MRI.

260 Circ Arrhythmia Electrophysiol June 2009

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

Figure 2. A, Original temperature recordings of a carbon and a metallic EP catheter. Catheters were placed in various configurationsinside the described gel phantom and MRI was performed to induce heating by RF pulses. Temperature increase was measured usinga fluoroptic thermometry system. Worst-case heating and heating at low RF transmit power is shown while running a pulse sequencewith a total acquisition time of 5 minutes. Black line indicates temperature evolution at the catheter tip; blue line, catheter middle; greenline, reference; red dots, start of pulse sequence and maximum temperature during MRI (averaged over 10 seconds). During MRI usingthe pulse sequences adapted for low RF transmit power (as described in the Results section), the carbon and the metallic cathetersremained at room temperature. B, Dependency of catheter tip heating on RF power applied to the tissue (SAR) in 2 representative con-figurations. Left graph, Conventional catheter (EPT blazer); right graph, carbon catheter. Solid line indicates half-circle catheter configu-ration inside the phantom liquid; dotted line, straight configuration. C, Standard versus carbon catheter heating in various configura-tions at a SAR of 2.1 W/kg. Mean heating, SD, and maximum heating (dots with numeric data) at the catheter tip, 200 mm behind thetip, and at the point where the catheter crossed from air to gel are given for 10 configurations each as described in the Methodssection. *P�0.05.

Nordbeck et al Hard- and Software Advances for MRI-Guided Ablation Therapy 261

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

carbon catheters, whereas the middle part of the cathetersshowed no significant heating (Figure 2C).

Comprehensive safety testing applying the low-SAR im-aging sequences did not reveal a significant temperatureincrease of the carbon catheters in any tested configuration(room temperature, 17�1.3°C; carbon catheter tip,17.3�0.7°C; n�12; P�NS).

Imaging of Carbon LeadsCatheter tracking in a 1.5-T scanner is displayed using agel phantom (Figure 3A) and in vivo in a pig heart (Figure3B). In the phantom, the 2 carbon catheters are clearlyvisible with no image artifacts in the surrounding area. Incontrast, the conventional metallic EP catheter causedlarge signal voids.

With MR-guided catheter placement, we targeted theright ventricular apex from a jugular access (Figure 3B).Images were acquired in end expiration using a breath-holdtechnique. The catheter did not remain in the imagingplane throughout the entire procedure but could be easilyrelocated with gradient echo or TrueFISP imaging. Passive

visualization of the steerable carbon catheter was possibleby the signal void and a small susceptibility artifact aroundthe catheter (2.5 mm, Figure 3B and Figure 5A). Again,metallic catheters caused large image voids in the adjacentarea, thus hampering precise tracking of their location.

In Vivo Animal StudiesFor diagnostic EP studies including programmed stimula-tion, the catheter tip was guided to different anatomicregions including the right atrium (lateral free wall), rightventricular apex, the His bundle region, and coronarysinus. Once the catheter tip had entered the atrium, anaverage of 4 imaging planes had to be acquired tosufficiently identify the tip using TrueFISP pulse se-quences as described in the Methods section. Subse-quently, the steerable catheter was actively directed to therespective anatomic sites. For accurate placement in theright atrium, right ventricle, and coronary sinus, 3 addi-tional real-time planes had to be used. Although identifi-cation of the His bundle ECG was possible and reproduc-

Figure 4. A, High-fidelity IEGM in His position. Detection ofthe His bundle ECG was possible with the carbon catheterunder MRI guidance. Arrows indicate His deflection. B, Pac-ing maneuver in the MR scanner. The carbon catheter wasplaced at the apex of the right ventricle during sinus rhythmat 89/min, then ventricular pacing was performed at a rate of100/min. C, Artifacts during MR scanning on intracardiacECG recordings.

Figure 3. Visualization of catheters in MRI and demonstrationof image void in vitro (FLASH imaging sequence). A, Two car-bon catheters and a conventional metallic EP catheter wereplaced in a gel-casted block next to a pig heart. The metalliccatheter caused a large area of image distortion. The carboncatheters were clearly localized with no surrounding imagevoid. The outer diameter of the carbon and the traditionalcatheter are 7F (2.33 mm). Resolution of the carbon catheterin the MR images was 2.3 to 2.5 mm. Determination of theresolution of the conventional catheter was hampered bymassive image void in the adjacent area. Signal loss wasminimum 9 mm (in vivo) to maximum 110 mm (in vitro). B,Left, Tip of carbon catheter at the right ventricular free wall.Right, Metallic EP catheter tip in the right atrium with a largesurrounding area of signal distortion.

262 Circ Arrhythmia Electrophysiol June 2009

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

ible (Figure 4A) a median of 4 different imaging planeswas necessary to localize the tip at the His region. Theaverage time to position the catheter at the targeted site andsensing and pacing thresholds are given in Table 1.

Pacing and Sensing QualitiesThe carbon catheter was capable of pacing and sensing and istherefore usable as a diagnostic EP catheter. The high-resolution IEGM was acquired in the control room outside thescanner via a shielded filter box and recorded with automateddata acquisition software (Figure 4A).

For cardiac pacing, an external pacer was connected to thecarbon catheter. The basal heart rate of the pigs was 50 to 100bpm, and pacing was performed in the VVI mode above thebasal heart rate (Figure 4B). R-wave amplitudes and capturethresholds are shown in detail in Table 2.

Acquiring unfiltered IEGMs via the carbon catheters dur-ing imaging picked up electromagnetic interference (Figure4C), whereas filtered intracardiac recordings were of highquality (Figure 4A).

RF AblationOne anatomic region was chosen in each animal for targetedablation as described. Right atrial sites (lateral wall and AVnodal region) were ablated in 1 animal each. Ablation in thecoronary sinus was performed in 2 animals. Ablation at the

apex of the right ventricle was performed in 4 animals(ventricular fibrillation occurred seconds after RF energyapplication in 1 animal). Figure 5A demonstrates targeting ofthe AV nodal area. The effect of RF ablation in the AVnodal/His area is demonstrated in Figure 5B. There wasdissociation of the atrial and ventricular rhythm indicatingsuccessful AV node modulation.

A detailed description of the ablation protocol is given inthe Methods section and in Table 2 (including RF energydelivery, time). The primary end point, significant voltagereduction of the respective signal in the IEGM tracings, wasachieved in each case. Average size of the ablation area was31.4�10 mm2 in the pathological specimen and 36.2�9 mm2

in MRI measurements (P�NS).Simultaneous MRI and RF ablation causes electromag-

netic interference resulting not only in IEGM disturbances(Figure 4C) but also large imaging artifacts. To overcomethis problem, RF filters had to be applied to suppress thesehigher harmonic signals and permit simultaneous ablationand imaging. The effect of these multistage filters onimage quality is shown in Figure 5C. The insert representsan image acquired during RF delivery without filtering; thelarge panel shows the same slice during RF delivery withfiltering.

Real-time MRI allows intraprocedural monitoring of ana-tomic structures in the direct vicinity of the targeted region(eg, esophagus, pericardial space), providing excellent infor-mation on possible complications. After proof of feasibilityfor ablation therapy, we intentionally caused perforation ofthe cardiac chambers in 2 animals to test whether the cathetertip could be tracked immediately in the pericardial space andif the development of pericardial effusion could be monitoredin the acute setting (Figure 6).

DiscussionRadiofrequency and pulsed gradient fields used for MRIinduce currents in the body.5–7,11–15 Such currents tend to beconcentrated in elongated, electrically conductive artificialstructures. Even severe burns have been reported.16 Recent

Table 1. Average Time to Position the Catheter at the TargetedSite and Sensing and Pacing Thresholds

RightAtrium

RightVentricle His

CoronarySinus

Time, s 134 � 39 334 � 51 291 � 41 147 � 77

Sensing, mV 3.5 � 1.7 13.1 � 5.6 0.9 � 0.4 3.9 � 2.0

Pacing, V/0.5 ms 0.7 � 0.4 0.7 � 0.5 0.9 � 0.5 1.7 � 0.8V

Time required to accurately guide the tip of the steerable, passive trackingcatheter to the desired region is given in seconds. All 4 regions were addressedin each animal. Pacing and sensing thresholds were achieved after MRIconfirmed correct anatomic position and are given as average values from 8animals.

Table 2. Pacing, R-Wave Amplitudes, and Capture Thresholds

Animal No. Scanner

Sensing/Pacing

Ablation Region Ablation Details ECG OutcomeBefore Ablation After Ablation

0 Siemens Vision – – V 60 s/75°C/23 W Asystole

1 Siemens Vision 12 mV/0.5@0.5 ms 2.3 mV/4.8@0.5 ms* V 180 s/75°C/65 W SR

2 Siemens Vision 19 mV/0.7@0.5 ms 0.7 mV/exit block* V 60 s/71°C/65 W SR

3 Siemens Avanto 9 mV/0.9@0.5 ms 3.5 mV/5.2@0.5 ms* V 60 s/72°C/48 W SR

4 Siemens Avanto 3 mV/0.4@0.5 ms 0.4 mV/3.5@0.5 ms* A 27 s/58°C/23 W AVB III

5 Siemens Avanto 4 mV/1.5@0.5 ms 1 mV/6.7@0.5 ms* CS 180 s/93°C/3 W SR

6 Philips Intera 12 mV/0.8@0.5 ms VF V 11 s/68°C/36 W VF

7 Philips Intera 3 mV/0.7@0.5 ms 0.5 mV/6.1@0.5 ms* A 60 s/75°C/33 W SR

8 Philips Intera 8 mV/0.7@0.5 ms 0.3 mV/5.1@0.5 ms* CS 120 s/73°C/11 W SR

Animal numbers are chronological. Animal 0 was killed before experiments to establish setup. For ablation details, the first number gives ablation times in seconds,the second number gives maximum temperature in degrees Celsius, and the third number gives maximum energy delivery in Watts. V indicates right ventricle; A,right atrium; CS, coronary sinus; SR, sinus rhythm; AVB III, 3rd-degree AV block; VF, ventricular fibrillation.

*Significant worsening of pacing and sensing threshold after ablation (P�0.05).

Nordbeck et al Hard- and Software Advances for MRI-Guided Ablation Therapy 263

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

reports, however, demonstrated feasibility of new develop-ments in the field of interventional EP in the MRI environ-ment.10,17–18 The current work further extended these studies,introducing a setup for MR-guided interventional EP andablation therapy including MR conditional carbon fiber–based catheters.

Safety TestingBecause of the large number of factors (eg, device length,diameter, material compounds, and orientation within the MRscanner) that contribute to the amount of heating in MRI,meeting the requirements for MR safety with elongatedconductive devices is extremely problematic. Therefore, thisstudy aimed to establish a scenario for MR-guided EPprocedures and ablation therapies that met the requirementsin a clinical surrounding. To this end, several methods werecombined that all decrease the risk for MRI-related heat-ing. In particular, the impact of MR imaging modalitieswith low SAR on heating of the catheters was assessed, asit was shown that heating is directly proportional to the RFpower applied by the MR scanner for a specific configu-

ration (Figure 2B).9 It is important to note that the carboncatheter did not display temperature changes during theclinically relevant scanning procedures and therefore isconsistent with the aim of being “MR conditional” in theexperimental setup described. We also assessed the possi-bility of tracking the catheter tip in the heart with low SARpulse sequences. Using the tested MRI sequences orsimilar “low SAR” sequences, it was possible to reproduc-ibly identify and guide the catheter to the desired targetregion with high precision.

ImagingAfter in vitro investigations on MRI safety of the carbonleads, they were tested for EP purposes. Typically, methodsfor device localization are characterized as either passive oractive catheter tracking. In passive techniques, materials arechosen so that the device is visible in the MR image. Activevisualization techniques use catheters that provide strongsignals10; however, extensive device modifications arerequired. It is demonstrated here that carbon catheters aresufficiently detectable in MRI without the need for active

Figure 5. A, Guiding of catheter to target sites (AV node). The first panel demonstrates the carbon catheter in a jugular sheath closethe right atrium. The second frame shows the tip catheter orientated to the lateral right atrial wall. The third panel shows the cathetertip close to the atrial septum. The His recording (Figure 3A) confirmed proximity to the AV node. Circles indicate catheter tip. B, ECGafter RF application in the AV nodal area. The ablation catheter was located in the atrium and indicated atrial activation. Ventricularactivation is shown in the leads I-III. Note the AV dissociation with accelerated ventricular rhythm after ablation. C, Artifacts of RF abla-tion on imaging without external filter box (insert) and with filter box (large panel). Arrow indicates catheter tip.

264 Circ Arrhythmia Electrophysiol June 2009

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

visualization techniques. Because the catheter componentsgive no MR signal, the signal void of the relatively thickcatheter is clearly visible in the image. At the same time,carbon catheters do not cause image voids in the surround-ing area, provided that sufficient imaging techniques areused.

Diagnostic EPRight atrial and ventricular sites and the coronary sinuswere successfully targeted in living animals using steer-able catheters with real-time MRI pulse sequences. Thehigh-resolution images of cardiac anatomy considerablyimproved targeting and accurate lesion placement, whereasstandard x-ray fluoroscopy in contrast generally deliverspoor tissue contrast. The pacing threshold did not changeduring MRI and was identical to thresholds observed inclinical practice. With the use of differential filter system,the IEGM could be received without electromagneticinterference.

RF AblationRecent developments in the area of interventional electro-physiology demonstrated the feasibility of electroanatomic

mapping and diagnostic EP studies in the MRI environ-ment.10,18 We now added proof that RF ablation is possible indifferent MR scanners using carbon catheters and specificsafety issues can be minimized provided adequate precau-tions. Ventricular fibrillation occurred in 1 animal duringablation of the right free ventricular wall. However, thisadverse event is a well-known complication of RF ablation atventricular targets and was not caused by the MRI environ-ment. We targeted anatomic regions that are regularly ad-dressed in clinical practice and demonstrated that imaging thecatheter tip and tracking the IEGM is possible during theentire procedures. Time, temperature, and RF delivery forablation therapy with the carbon catheters to achieve myo-cardial lesions were similar to those by conventional EPcatheters. This approach might help to improve safety andefficacy of RF ablation, especially for ablation of complexarrhythmias, in which knowledge of the neighboring anatomyand possible gaps for conduction is crucial for success andsafety.

LimitationsAlthough the current study presents major advances inMR-guided interventional EP, there are still limitations re-

Figure 6. Real-time demonstration of complications during ablation procedures in MRI. A, Carbon ablation catheter was directed to theright ventricle via the jugular vein. During ablation, the ablation catheter was intentionally pushed further forward, causing perforation ofthe interventricular septum. The catheter tip was located subepicardial at the apex of the left ventricle, which could be directly moni-tored by MRI. The small pericardial separation indicates pericardial effusion. Detection of this complication was in the first FLASH cinesequence 30 seconds after ablation. B, In situ view of the left ventricle. The thermal lesion was located subepicardial on the left ventri-cle after perforation from the right ventricle. C, Demonstration of the tunnel from right to left ventricle at gross examination. D, Demon-stration of developing pericardial effusion after RF delivery. In this case, ablation was performed at the ostium of the coronary sinus. 1minute after cessation of ablation, beginning pericardial effusion was detected, slowly expanding afterward. The panels show pericar-dial effusion over time in a short-axis view 3 and 5 minutes after ablation.

Nordbeck et al Hard- and Software Advances for MRI-Guided Ablation Therapy 265

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

garding investigations in certain patients. This issue hasrecently been reviewed in an American Heart Associationscientific statement on the safety of MRI in patients withcardiovascular devices.19 Also, image quality is compromisedbecause of the necessity of low SAR protocols for safetypurposes. Therefore, a number of technical challenges mustbe resolved before its widespread clinical use can be realized.There is a need for specialized hardware for MRI-guidedinterventional electrophysiology studies, and further ad-vances in imaging techniques must be established to useMR-guided electrophysiological interventions to full capac-ity. As an example, MR-based visualization of scar morphol-ogy was recently shown to contribute to preproceduralplanning for interventional EP.20

Successful assessment of ablation sites using enhanced andnonenhanced MRI protocols was recently described.3,4 Al-though this study was not intended to investigate lesionmapping strategies, we preliminarily investigated the possi-bilities to gain intraprocedural information of the lesiondevelopment. In this context, it is important to mention thatassessment of the lesion size in the MRI was performedshortly after ablation to test for the clinical relevance of thisprocedure, whereas pathological examination followed sev-eral hours later. Clearly, more studies are warranted furtherextending the possibilities and clinical implications of theseapproaches.

Conclusion and Clinical ImplicationsThe current study demonstrates the feasibility of interven-tional EP studies and targeted RF ablation under real-timeMRI guidance using novel carbon catheters. Furthermore,it gives proof of improved imaging qualities and safetyproperties for these catheters with specific short low-SARimaging sequences. Although the approach described hereis applicable to all cardiac arrhythmias, it might beparticularly well suited for more complex arrhythmias thatrequire the accurate placement of multiple, linearly ar-ranged lesions (eg, atrial flutter and fibrillation, ventriculartachycardia). In addition to improved anatomic targeting ofcritical focal sites, the ability to directly visualize thespatial extent of lesions with high spatial resolution mayhelp to facilitate the placement of linear transmural lesionsand allow for real-time detection (and possibly prevention)of complications. These advances might help to reduce thenumber of lesions required for conduction block, proce-dure time, and potential procedural risks such as perfora-tion, all without ionizing radiation or iodized contrastagents in the future.

Sources of FundingThis work was supported by the Bayerische Forschungsstiftung(Project “Entwicklung MR-kompatibler Schrittmacherelektroden”)and the Deutsche Forschungsgemeinschaft (SFB 688, B2).

DisclosuresDrs Ritter and Bauer serve as scientific advisors for Biotronik, andDr Jakob serves as scientific advisor for Siemens. Drs Maxfield,Wurtz, and Geistert are employees of Biotronik.

References1. Cappato R, Calkins H, Chen SA, Davies W, Iesaka Y, Kalman J, Kim

YH, Klein G, Packer D, Skanes A. Worldwide survey on the methods,efficacy, and safety of catheter ablation for human atrial fibrillation.Circulation. 2005;111:1100–1105.

2. Lardo AC, McVeigh ER, Jumrussirikul P, Berger RD, Calkins H, Lima J,Halperin HR. Visualization and temporal/spatial characterization ofcardiac radiofrequency ablation lesions using magnetic resonanceimaging. Circulation. 2000;102:698–705.

3. Dickfeld T, Kato R, Zviman M, Lai S, Meininger G, Lardo AC,Roguin A, Blumke D, Berger R, Calkins H, Halperin H. Character-ization of radiofrequency ablation lesions with gadolinium-enhancedcardiovascular magnetic resonance imaging. J Am Coll Cardiol. 2006;47:370 –378.

4. Dickfeld T, Kato R, Zviman M, Nazarian S, Dong J, Ashikaga H, LardoAC, Berger RD, Calkins H, Halperin H. Characterization of acute andsubacute radiofrequency ablation lesions with nonenhanced magneticresonance imaging. Heart Rhythm. 2007;4:208–214.

5. Duru F, Luechinger R, Scheidegger MB, Luscher TF, Boesiger P,Candinas R. Pacing in magnetic resonance imaging environment: clinicaland technical considerations on compatibility. Eur Heart J. 2001;22:113–124.

6. Luechinger R, Zeijlemaker VA, Boesiger P, Duru F. Potential harmfuleffects of magnetic resonance imaging in pacemaker patients should notbe underestimated. Europace. 2006;8:389–390.

7. Luechinger R, Zeijlemaker VA, Pedersen EM, Mortensen P, Falk E,Duru F, Candinas R, Boesiger P. In vivo heating of pacemaker leadsduring magnetic resonance imaging. Eur Heart J. 2005;26:376 –383.

8. Aube C, Schmidt D, Brieger J, Schenk M, Helmberger T, Koenig CW,Schick F, Claussen CD, Pereira PL. Magnetic resonance imaging char-acteristics of six radiofrequency electrodes in a phantom study. J VascInterv Radiol. 2004;15:385–392.

9. Nordbeck P, Fidler F, Weiss I, Warmuth M, Friedrich MT, Ehses P,Geistert W, Jakob PM, Ladd ME, Quick HH, Bauer WR. Spatial distri-bution of RF-induced E-fields and implant heating in MRI. Magn ResonMed. 2008;60:312–319.

10. Nazarian S, Kolandaivelu A, Zviman MM, Meininger GR, Kato R, SusilRC, Roguin A, Dickfeld TL, Ashikaga H, Calkins H, Berger RD,Bluemke DA, Lardo AC, Halperin HR. Feasibility of real-time magneticresonance imaging for catheter guidance in electrophysiology studies.Circulation. 2008;118:223–229.

11. Gimbel JR, Kanal E. Can patients with implantable pacemakers safelyundergo magnetic resonance imaging? J Am Coll Cardiol. 2004;43:1325–1327.

12. Kanal E, Borgstede JP, Barkovich AJ, Bell C, Bradley WG, Etheridge S,Felmlee JP, Froelich JW, Hayden J, Kaminski EM, Lester JW Jr, ScoumisEA, Zaremba LA, Zinninger MD. American College of Radiology WhitePaper on MR Safety: 2004 update and revisions. AJR Am J Roentgenol.2004;182:1111–1114.

13. Nitz WR, Oppelt A, Renz W, Manke C, Lenhart M, Link J. On theheating of linear conductive structures as guide wires and catheters ininterventional MRI. J Magn Reson Imaging. 2001;13:105–114.

14. Shellock FG. MR imaging and electronically activated devices.Radiology. 2001;219:294–295.

15. Sommer T, Vahlhaus C, Lauck G, von Smekal A, Reinke M, Hofer U,Block W, Traber F, Schneider C, Gieseke J, Jung W, Schild H. MRimaging and cardiac pacemakers: in-vitro evaluation and in-vivostudies in 51 patients at 0.5 T. Radiology. 2000;215:869 – 879.

16. Kugel H, Bremer C, Puschel M, Fischbach R, Lenzen H, TombachB, Van Aken H, Heindel W. Hazardous situation in the MRbore: induction in ECG leads causes fire. Eur Radiol. 2003;13:690 – 694.

17. Greatbatch W, Miller V, Shellock FG. Magnetic resonance safety testingof a newly-developed fiber-optic cardiac pacing lead. J Magn ResonImaging. 2002;16:97–103.

18. Dukkipati SR, Mallozzi R, Schmidt EJ, Holmvang G, d’Avila A, GuhdeR, Darrow RD, Slavin G, Fung M, Malchano Z, Kampa G, Dando JD,McPherson C, Foo TK, Ruskin JN, Dumoulin CL, Reddy VY. Electro-anatomic mapping of the left ventricle in a porcine model of chronicmyocardial infarction with magnetic resonance-based catheter tracking.Circulation. 2008;118:853–862.

19. Levine GN, Gomes AS, Arai AE, Bluemke DA, Flamm SD, KanalE, Manning WJ, Martin ET, Smith JM, Wilke N, Shellock FS.

266 Circ Arrhythmia Electrophysiol June 2009

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from

Safety of magnetic resonance imaging in patients with cardiovasculardevices: An American Heart Association Scientific Statement Fromthe Committee on Diagnostic and Interventional Cardiac Catheteriza-tion, Council on Clinical Cardiology, and the Council on Cardio-vascular Radiology and Intervention. Circulation. 2007;116:2878 –2891.

20. Ashikaga H, Sasano T, Dong J, Zviman MM, Evers R, Hopenfeld B,Castro V, Helm RH, Dickfeld T, Nazarian S, Donahue JK, Berger RD,Calkins H, Abraham MR, Marban E, Lardo AC, McVeigh ER, HalperinHR. Magnetic resonance-based anatomical analysis of scar-related ven-tricular tachycardia: implications for catheter ablation. Circ Res.2007;101:939–947.

CLINICAL PERSPECTIVEMRI can provide detailed views of cardiac morphology and function. Real-time MR imaging during electrophysiological(EP) studies has the potential to improve visualization of anatomic detail and show ablation lesions but requiresMR-compatible equipment, which is a substantial challenge. Recent work has shown feasibility of performing EP studiesand electroanatomic mapping in MR scanners. The current study extends previous findings. A specialized EP setup thatincludes MR conditional catheters based on carbon technology is presented. This new technology allows for diagnostic EPstudies and RF ablation therapy in the MR environment. We demonstrated successful targeting of different anatomicregions in the heart, with subsequent ablation in selected regions. It is hoped that further development of this technologywill make real-time MR-guided EP procedures a clinical reality to facilitate complex EP interventions.

Nordbeck et al Hard- and Software Advances for MRI-Guided Ablation Therapy 267

at Universitaet Wuerzburg on October 21, 2009 circep.ahajournals.orgDownloaded from